M1021 Endoscopy Unit Quality Indicators for Colonoscopy: the Canadian Association of Gastroenterology (cag) Endoscopy Quality Initiative (EQI) Practice Audit Project David Armstrong, Roger Hollingworth, Donald G. MacIntosh, Ying Chen, Ronald Bridges, Sandra Daniels, Stuart Gittens, Paul Sinclair, Joanne Cabrera, Catherine Dube Background: Endoscopy units' delivery of timely service, consistent with published guidelines (Can J Gastro 2006;20:411) depends, partly, on efficient use of available resources. However, there is no mechanism to allow direct widespread monitoring of systemic quality indicators, such as booking appropriateness and bowel preparation adequacy in Canadian endoscopy units. Aim: To determine endoscopy unit quality indicators in the context of a national colonoscopy practice audit program. Methods: MDs at 13 centres collected data, in real time, on colonoscopies performed over periods of at least 2 weeks using data collection software (ReForm XT, Goanyware Software, Tulsa, OK) on a smartphone (Treo 650, Palm Inc, Mississauga, ON) with prompt download to a secure website and presentation on a secure website (ECD solutions, Atlanta, GA) for participants' review. Reasons for colonoscopy were classified as ‘Investigation of abnormality' (INV), ‘Screening' (SCR) and ‘Surveillance' (SUR). Bowel preparation quality was rated with the Ottawa scale (Scores - Excellent: 0-4; Poor: 11-14). Results : Since February 2008, 45 endoscopists (34 GIs & 11 surgeons) reported on 822 colonoscopies. Bowel preparation was: excellent: 73.9% and poor: 2.7%. The table shows the proportions of patients (a) colonoscoped at different intervals since their last colonoscopy and (b) in each age group at the time of colonoscopy. Summary: Colonoscopy intervals seem appropriate and most patients have good bowel preparation. However, for SCR patients, 20% are ≤50 yrs old, >10% have had a colonoscopy within <5 yrs and 8% are >70 yrs old. Conclusions: Point-of-care data collection on reasons for colonoscopy and bowel preparation quality facilitates: 1) practice audit by physicians and endoscopy units, and 2) improved resource utilization in the delivery of colonoscopy services. Percentages [95% CI] of patients with respect to interval since last colonoscopy and age '-' No patients in this group. M1022 African-American Race Does Not Affect Compliance with Recommended Colonoscopy Maria C. Hatara, Jayme Tishon, Marie L. Borum INTRODUCTION Colorectal cancer causes significant morbidity and mortality in the United States. African-Americans are disproportionately affected by this malignancy when compared to other ethnic groups. Screening rates in African-Americans is lower than that of other ethnic groups. It has been reported that African-Americans are less compliant with colorectal cancer screening recommendations. This study evaluated the compliance of African-Amer- icans who had access to health care with recommended colonoscopy in an urban university gastroenterology practice. METHODS A retrospective medical record review of consecutive patients referred for colonoscopy during a 6-month period was performed. There were no exclusion criteria. Patient race, age and gender were obtained. A database was created using Microsoft Excel. Identifying factors were eliminated to ensure patient confidentiality. Multivariate analysis was performed using Fisher's exact test, Chi-square and relative risk ratios. The study was approved by the institutional review board. RESULTS The medical records of 985 patients referred for colonoscopy were reviewed. There were 577 women and 408 men. Two hundred forty-nine patients (25%) were white, 407 (41%) were African- American, 37 (4%) were Hispanic and 15 (2%) were Asian. Two hundred seventy-five (28%) patients were of unknown race / ethnicity. Three hundred one (30.5%) were compliant with A-333 AGA Abstracts the recommended colonoscopy. There was no significant difference in the rate at which patients complied with colonoscopy based upon age or gender. Multivariate analysis revealed that there was no significant difference in the rate at which African-Americans complied with recommended colonoscopy when compared to other ethnic groups. (RR=0.97; 95% CI=0.76-1.22) CONCLUSION This study revealed that compliance with colonoscopy is suboptimal in all patients. There was no significant difference in the rate at which African- Americans complied with recommended colonoscopy compared to other ethnic groups. While African-Americans may have a lower colorectal cancer screening rate, it is not a result of disproportionate noncompliance with recommendations when there is access to medical care. It is critical that factors which result in decreased compliance in all patients are identified. Increased efforts are needed to improve colorectal cancer screening in those patients who have not sought or do not have regular access to health care. M1023 A Call to Screen - Clinical Correlates and Consequences of Colorectal Cancer Among Kidney Transplant Candidates and Recipients: Should We Revisit Our Screening Practices? Jennifer L. LaBundy, Krista Lentine, Huiling Xiao, Patrick M. Ercole, Mark Schnitzler, Charlene M. Prather Background: No specialized guidelines exist for colorectal cancer (CRC) screening in renal transplant patients. Recent data suggest an increased risk of CRC in these patients compared to the general population. We examined the incidence and mortality implications of CRC in renal transplant candidates on the waitlist, and renal transplant recipients. Methods: Registry data collected by the United States Renal Data System was used in this retrospective cohort study. 45,238 patients possessing Medicare 1 year prior to transplant, with no prior diagnosis of CRC, underwent renal transplant in 1995 to 2004. We ascertained CRC diagnoses from billing records and estimated incidence of first events by the product-limit method. Stepwise Cox regression was used to identify independent correlates of CRC and to examine these events as time-dependent mortality predictors. Results: For transplant recipients, 1 and 3 year cumulative incidences of new onset CRC were 0.19% and 0.58%, respectively, and 0.021% and 0.90% in the waitlist group. When stratified by clinical characteristics, age significantly correlated with CRC incidence in both the transplant and waitlist groups. In the multi-variate model of CRC after transplant, risk compared to the youngest recipients age 18-30 increased with age, particularly after 40 years (See Table). In transplanted patients, overweight BMI (≥25 to <30) was associated with increased risk of CRC compared to all other BMI groups. CRC, modeled as a time-dependent variable, was associated with a 52% reduction in the likelihood of progressing to transplant after listing (aHR 0.484; 95% CI 0.29-0.80). CRC was a strong predictor of death in the transplant group (aHR 4.07, 95% CI 3.01-5.49) and waitlist group (aHR 1.82, 95% C.I. 1.21-2.75) compared to those in the cohort who remained without CRC. Conclusion: CRC impairs access to transplant among candidates and confers a high mortality risk. Until further studies are available outlining optimal screening guidelines for pre and post-transplant patients, CRC screening should be strongly considered in all patients ≥ 40 years of age. Even patients mildly overweight should be considered at increased risk for CRC and appropriately screened. 1 Relative risks of CRC associated with these baseline factors, age group 18-29 as reference M1024 Incidence and Clinical Characteristics of Colorectal Cancer in a Safety-Net Hospital System: Implications for U.S. National Healthcare Policy Jessica Shah, Marisa Valdes, Don C. Rockey Background: Safety-net health systems provide care to a large proportion of United States' under- and uninsured population. We aimed to understand the incidence and clinical characteristics of patients with colorectal cancer (CRC) in a safety-net health system. Methods: We collected demographic and clinical data on the 318 patients seen with CRC at Parkland Health & Hospital System (PHHS, Dallas, TX), a safety-net health system, from 2005 to 2007. Patients were identified by the hospital's cancer registry. We tabulated and compared the data to representative data over the same time frame from other regional health systems. Chi Square statistic was used for comparison of groups; a p-value of <0.05 was considered significant. Results: The proportion of African Americans and Hispanics with CRC seen at PHHS was higher compared to other regional health centers [46% vs. 20%, (p<0.001) and 21% vs. 8%, (p<0.001), respectively]. PHHS patients were more likely to have metastatic disease compared to the other centers (42% vs. 18%, p<0.001). 25% (80/318) of PHHS patients had prior CRC screening; those with prior screening were more likely to have Stage 1 or Stage 2 disease (56% vs. 33%, OR 2.54, 95% CI: 1.48-4.38). Interestingly, 89% (285/ 318) of patients had symptoms (blood in stool, abdominal pain, stool changes or weight loss) or iron deficiency anemia prior to diagnosis. Only 4% of PHHS patients had private insurance, 21% had Medicare, and 15% had Medicaid; the remaining 60% were uninsured or were on a local tax-supported assistance program. Patients with Medicare or private insurance had lower rates of Stage 3 & 4 disease (25% vs. 68%, p<0.001). We further analyzed the data to establish each patient's portal of entry within PHHS, leading to the initial diagnosis of CRC (Table 1); 75% of patients were diagnosed by an outside medical facility or at the PHHS emergency room and subsequently sought care at PHHS. Conclusions: AGA Abstracts